Business runs on data. Decisions made without reliable, properly gathered data are, at best, educated guesses.
At worst, decisions made without accurate data can put a company out of business.
As important as the data’s source are the methods and procedures used to optimize that data. “Data optimization” is how data are used and aligned to most efficiently meet the goals of the organization.
That’s where data analytics come into play.
Part of the job of data or business analysts is to ensure that an organization’s leaders have the information they need to make informed decisions. Executives want to know that the data used to make decisions is gathered in a reliable, systematic way.
Effective data collection is an integral tool for a company to build a competitive advantage, but it could be even more vital than that: A company’s survival might depend on it.
If you are in the information technology field, you no doubt have heard the term “garbage in, garbage out.” You might even have seen its acronym, GIGO, and recognized it as a verbal alarm bell for shoddy results.
Preventing “garbage-in, garbage-out” syndrome from dragging your organization into a fiscal morass begins with thoughtful preparation. One key to effective data gathering is to understand the differences between quantitative and qualitative data, and how they are compiled and used (separately or in tandem).
Quantitative data, or research that produces measurable, numerical results, consist of four basic scales of measurement:
- Nominal: A classification of data in which subjects are categorized by groups for contrast or comparison using codes such as colors, letters or numbers; examples include students vs. non-students, demographic information and age.
- Ordinal: In this classification, results are ranked lowest to highest and are displayed in comparison to one another. Examples include annual rainfall data, month-to-month raw material costs and amounts of a particular product consumed over an assigned period of time by particular subject groups.
- Interval: This is a measurement of distance, variable physical qualities (such as temperature) and/or elapsed time. An example is a thermometer, which measures current temperature at a particular time and can be used to discover variations in order to derive ordinal data.
- Ratio: This scale combines nominal, ordinal and interval scales to produce a data-driven “big picture” to help analysts understand relationships among different pieces of data. Measurements of time are an example of the ratio scale in effect.
Qualitative data are measurements of subjective beliefs and opinions gathered through direct interviews, surveys, publicly presented statements of opinion and other gathering methods. While most qualitative data are gathered through similar methods, the purpose and approach to using this type of data can vary greatly.
Qualitative research approaches include:
- Ethnographic is the study of an entire culture or organization based on ethnic makeup, geography, shared political beliefs and other physical, emotional or historical traits.
- Phenomenological is based on personal experience and worldview, and seeks to discover why people interpret the world as they do.
- Field research is based on personal observation and, at times, participation among the population under study.
- Grounded theory seeks to discern unifying factors among different sets of observations and data.
Most organizations will need to make use of quantitative and qualitative data in order to determine whether a strategy is sound, or whether adjustments must be made from quarter to quarter or year to year.
Quantitative data is used primarily to create usable statistics and uncover patterns. Qualitative data is often used to determine answers to open-ended questions or inquiries about an opinion. This is an important distinction, because data analysts often are charged with figuring out how best to gather data effectively, and what kind of data must be gathered.
That means developing a clear, well-thought-out plan based on the needs and goals of your organization. A good way to begin to determine how effectively to gather and optimize your organization’s data is to answer these questions posed on the website Principles for Digital Development:
- What question do you seek to answer?
- What outcomes (measurable results) do you expect or want to gather?
- What indicators will you define as a way to measure outcomes?
The answers to those questions should align with goals that are clearly established by the stakeholders in your organization, especially the executives whose decisions will be influenced by the data. As those questions are answered, an analyst might be expected to provide further insight in an effort to narrow the focus of the research.
In order to do this, it is important to answer these three questions:
- What kind of information do you need?
- What information can be collected easily, with minimal effort and cost?
- What kind of information is most relevant to the organization?
Proper care must be taken to be sure that the answers to all of these developmental questions lead to the gathering of relevant data. Only then will your organization avoid the perils of “garbage-in, garbage-out” syndrome.